A Governance Framework for Operationalising Responsible AI Use in Students’ Research Projects: Illustrations from Viet Nam

Abstract

Generative artificial intelligence (AI) is increasingly used in students’ research projects for literature searching, idea development, drafting, language editing, information processing, and research planning. These uses can support research learning, but they also raise academic integrity concerns, including unverifiable sources, fabricated references, undeclared AI assistance, over-reliance on AI-generated outputs, unclear authorship, and unfair assessment. This conceptual paper argues that responsible AI use in students’ research projects cannot be secured through individual AI literacy or general guidance alone. Higher education institutions also need governance mechanisms that make AI use transparent, verifiable, assessable, and accountable. Using targeted literature identification and structured conceptual synthesis, the paper maps recurring governance problems to normative principles, institutional requirements, and operational mechanisms. It proposes a five-component framework corresponding to five governance functions: policy (boundary setting), transparency (traceability), assessment (evaluability), audit (assurance), and guidance/support (capability building). Viet Nam is used as an illustrative emerging AI governance context, where national policy directions on digital competence and AI are developing but operational mechanisms for student research remain under-specified. The framework and associated tools are presented as adaptable design artefacts for future institutional testing rather than as validated interventions.